Replication Material for "Perspective Taking Through Partisan Eyes"

## 1 Information 

Authors: Don Casler and Dylan Groves
Date: July 12, 2022
Contact email: donald.casler@columbia.edu; dylanwgroves@gmail.com

### ### ### ###

# 2 File overview

# 2.1 Data files

The paper draws on three sources of data:
		
		(1) Mturk Sample (collected in two parts), referred to as "Wave 1"
			ptinir_data_wave1_raw.dta
			
		(2) Lucid Sample (collected in one part), referred to as "Wave 2"
			ptinir_data_wave2_raw.dta
			
		(3) Prolific Sample (collected in one part), referred to as "Wave 3"
			ptinir_data_wave3_raw.dta
		
		Note that the treatment and outcome measurements are separate for the 
		Prolific sample. 

# 2.2 Code files

Most of the data analysis was conducted in Stata. There are cleaning files, one for each of the three sources of data. Each cleaning file draws on a separate raw data file. These include:

		(1) "ptinir_replication_wave1_clean.do" for Wave 1
		
		(2) "ptinir_replication_wave2_clean.do" for Wave 2

		(3) "ptinir_replication_wave3_clean.do" for Wave 3

To begin, first move all extracted raw data and do files to the same folder and set the global using the following code:

		gl stata "[WORKING DIRECTORY]"

The following code and file appends the cleaned data from the three waves together:
		
		do "${stata}/ptinir_replication_merge.do"

The following code and file conducts analysis for all tables found in the paper. You can identify and run one table at a time or all tables together. 

		do "${stata}/ptinir_replication_analysis.do"

The figures that appear in the paper and appendix were created in R. There is one R script ("ptinir_replication_figures.R" that loads the master .dta file ("ptinir_pooled.dta") and generates the figures.

# 2.3 Software versions

All analyses in this article were carried out using either Stata 16 in Windows 11 on PC or R Studio Version 2022.02.3 and R version 4.2.0 for Mac.

# 2.4 Codebook

The codebook is included as a PDF and details the variable names used in the analysis.

# 3 Data source files 

The following files contain the raw data by experimental wave:

		(1) "ptinir_data_wave1_raw.dta" for Wave 1

		(2) "ptinir_data_wave2_raw.dta" for Wave 2

		(3) "ptinir_data_wave3_raw.dta" for Wave 3

The file "ptinir_pooled.dta" contains the fully merged data.

# 4 List of required packages

Note: all packages are available from CRAN as of 7/12/2022

(BayesTree, data.table, disco, deployer, estimate, FindIt, ggplot2, ggpubr, gif, plotrix, rgeos, rgdal, readstata13, RColorBrewer, sf, sp, tidylverse, tree, grid, magrittr, quanteda, stm)




